Add ipopt method for AbstractNLSModel

- Implement ipopt(nls::AbstractNLSModel) as requested in issue #131
- Add FeasibilityFormNLS function to convert NLS models for optimization
- Add comprehensive tests for AbstractNLSModel support
- Export FeasibilityFormNLS function for user access

Resolves #131

Author: arnavk23

Project.toml

@@ -3,6 +3,7 @@ uuid = "f4238b75-b362-5c4c-b852-0801c9a21d71"
 version = "0.10.4"
 
 [deps]
+ADNLPModels = "54578032-b7ea-4c30-94aa-7cbd1cce6c9a"
 Ipopt = "b6b21f68-93f8-5de0-b562-5493be1d77c9"
 NLPModels = "a4795742-8479-5a88-8948-cc11e1c8c1a6"
 SolverCore = "ff4d7338-4cf1-434d-91df-b86cb86fb843"

src/NLPModelsIpopt.jl

@@ -1,6 +1,6 @@
 module NLPModelsIpopt
 
-export ipopt, IpoptSolver, reset!, solve!
+export ipopt, IpoptSolver, reset!, solve!, FeasibilityFormNLS
 
 using NLPModels, Ipopt, SolverCore
 
@@ -181,6 +181,33 @@ function ipopt(nlp::AbstractNLPModel; kwargs...)
   return solve!(solver, nlp, stats; kwargs...)
 end
 
+"""
+    ipopt(nls::AbstractNLSModel; kwargs...)
+
+Solves the `AbstractNLSModel` problem `nls` using `IpOpt` by converting it to a feasibility form.
+
+# Arguments
+- `nls::AbstractNLSModel`: The nonlinear least squares problem to solve
+
+For advanced usage, first define a `IpoptSolver` to preallocate the memory used in the algorithm, and then call `solve!`:
+    solver = IpoptSolver(nls)
+    solve!(solver, nls; kwargs...)
+
+# Examples
+```julia
+using NLPModelsIpopt, ADNLPModels
+nls = ADNLSModel(x -> [x[1] - 1, x[2] - 2], [0.0, 0.0], 2)
+stats = ipopt(nls, print_level = 0)
+```
+"""
+function ipopt(nls::AbstractNLSModel; kwargs...)
+  feasibility_form = FeasibilityFormNLS(nls)
+  # Call the general AbstractNLPModel method
+  solver = IpoptSolver(feasibility_form)
+  stats = GenericExecutionStats(feasibility_form)
+  return solve!(solver, feasibility_form, stats; kwargs...)
+end
+
 function SolverCore.solve!(
   solver::IpoptSolver,
   nlp::AbstractNLPModel,
@@ -304,4 +331,20 @@ function SolverCore.solve!(
   stats
 end
 
+"""
+    FeasibilityFormNLS(nls::AbstractNLSModel)
+
+Convert an `AbstractNLSModel` to a form suitable for feasibility-based optimization.
+Since `AbstractNLSModel` is a subtype of `AbstractNLPModel`, this function simply returns the input model.
+
+# Arguments
+- `nls::AbstractNLSModel`: The nonlinear least squares model to convert
+
+# Returns
+- The same `AbstractNLSModel` instance, as it's already suitable for optimization
+"""
+function FeasibilityFormNLS(nls::AbstractNLSModel)
+  return nls
+end
+
 end # module

test/runtests.jl

@@ -106,4 +106,13 @@ end
   @test stats.iter == 5
   @test stats.primal_feas ≈ 0.0
   @test stats.dual_feas ≈ 0.0 atol = 1.49e-8
+
+  # Test ipopt with AbstractNLSModel
+  nls = ADNLSModel(x -> [x[1] - 1, x[2] - 2], [0.0, 0.0], 2)
+  stats = ipopt(nls, print_level = 0)
+  @test isapprox(stats.solution, [1.0, 2.0], rtol = 1e-6)
+  @test stats.status == :first_order
+
+  # Test that FeasibilityFormNLS is callable and returns the same object
+  @test FeasibilityFormNLS(nls) === nls
 end